" --- -i ....:_._ ....... as of Wide-Field-of-View going Longwave Radiation _rived From Nimbus 7 Earth "m Budget Data Setm vember 1985 to October 1987 T. Da]eBess and G. Louis Smith := 7 < 7 _ _ i_-imi_7.- ¸ :-- _=_- In m =. NASA Reference Publication 1261 1991 Atlas of Wide-Field-of-View Outgoing Longwave Radiation Derived From Nimbus 7 Earth Radiation Budget Data Set-- November 1985 to October 1987 T. Dale Bess and G. Louis Smith Langley Research Center Hampton, Virginia National Aeronautics and Space Administration Office of Management Scientific and Technical Information Program Introduction this atlas were derived with a deconvolution (i.e., a resolution enhancement) technique that represented The Earth's radiation balance is an important the WFOV monthly averaged OLR as an expression factor in climate change. One important parameter of spherical harmonic coefficients (Smith and Green in this balance is how the Earth's outgoing long-wave 1981). Tables of these coefficients and monthly radiation (OLR) is changing on a regional, zonal, averaged contour maps of OLR results for 2 years and global scale in the spatial domain and on a are included. monthly, annual, and interannual scale in the time domain. To focus on long-term climate processes, The results documented in this atlas are impor- one requires a long _ime series (>20 years) of high- tant for a number of reasons. One reason relates to quality measurements of OLR. the data, which are both broadband and WFOV. Be- cause the measurements are broadband, the ERB ra- For the past 20 years radiometers aboard satel- diometers offer some significant advantages over the lites have provided comprehensive measurements of instruments aboard the National Oceanic and At- OLR. In 1975, the first of two satellite experiments mospheric Administration (NOAA) operational-type using new concepts for measuring the Earth's radia- polar orbiting satellites. The NOAA instruments tion budget began. These new experiments were the measure upwelling radiation in the narrow spectral Earth radiation budget (ERB) experiments (Smith et regions (0.5 to 0.7 #m in the visible region and 10.5 al. 1977; Jaeobowitz et al. 1979, 1984) that flew on to 12.5 pm in the infrared region). Winston et al. the polar orbiting Nimbus 6 and Nimbus 7 satellites. (1979) published an atlas that documented 4 years The ERB experiments were launched into nearly of OLR results from the scanning radiometer on circular Sun-synchronous Earth orbits. The ERB in- the NOAA operational satellites. Janowiak et al. strument package included two Earth-viewing, wide- (1985) also published an atlas of OLR results derived field-of-view (WFOV) radiometers, one of which from NOAA operational satellite data that covered measured total (TOT) irradiance and the other of the time period from June 1974 to November 1983. which measured shortwave (SW) irradiance. The dif- The primary disadvantage of measurements made by ference between the two (TOT SW) is the OLR. NOAA operational satellites is that they are made in These WFOV radiometers viewed the entire Earth's narrow spectral regions that must then be empirically disk from the satellites, which had altitudes of almost corrected to estimate the broadband OLR. The pri- 1000 km and Equator crossing times near 12:00 a.m. mary advantage of these measurements is their very (noon) and 12:00 p.m. (midnight). The SW channel high spatial resolution compared with the WFOV ra- had a spectral range of 0.2 to 3.8 #m, and the total diometer, which is limited to large scales. channel measured the irradiance from 0.2 to 50+ #m. The WFOV instrument is well suited for measur- The first of the ERB experiments was' on the ing large-scale features since its field of view is such Nimbus 6 satellite and was launched in June 1975. that the instrument measures all incident radiation The second experiment was launched on the Nim- from horizon to horizon. Herein lies one of its advan- bus 7 satellite in November 1978. The combined tages because the measured radiation is integrated OLR record from the ERB instruments from Nim- ove r a broad variation of angles, and thus it is less bus 6 and Nimbus 7 spanned the time period from sensitive to directional models than the data mea- July 1975 to October 1987, with a 4-month gap (July sured with narrow-field-of-view radiometers. Also, 1978 to October 1978). Atlases of a 10-year time se- because of their mechanical simplicity, fixed-WFOV ries of monthIy averaged OLR data were published in radiometers typically have much greater longevity two NASA Reference Publications (Bess and Smith than scanning radiometers. 1987a, 1987b). Since that time, the next 2 years (mi- In addition to the data being broadband and nus May 1986) of OLR data for Nimbus 7 from the WFOV, the method of representing the data using Nimbus 7 Project have been received. This paper is spherical harmonics is important. The spherical har- an atlas of the 23 months of monthly averaged OLR monic coefficient data set for each month represents results from Nimbus 7 estimated at the top of the a condensation of the OLR field. Some of these co- atmosphere (denoted by TOA and defined herein to efficients have a physical interpretation attached to be at an altitude of 30 km) and spanning the time them. These coefficients may be analyzed individu- period from November 1985 to October 1987. ally or in combination to study different aspects of The purpose of this atlas is to document all the the radiation field. The Nimbus 7 results from this WFOV OLR results from the 2 years of Nimbus 7 atlas and the 10-year data set already published give operation in a form that facilitates analysis of the a 12-year time series of OLR that will be very useful Earth's radiation field. The results contained in for doing monthly, annual, and interannual studies of OLRon large spatial scales. The last 3 years of the in latitude and 5° or greater in longitude. This grid data set also give a 3-year overlap with the current system formed an igloo-type grid of near-equal-area Earth Radiation Budget Experiment (ERBE) data regions that were symmetrical about the Equator, set that will be used to intercompare the two data with 3 grid boxes between 85 ° and 90 ° latitude sets. increasing to 72 grid boxes near the equatorial region (Bess and Smith 1987a, 1987b). Data Processing and Analysis Some corrections and editing had to be made The 2 years of ERB WFOV irradiance data from to the ERB WFOV measurements before they were the Nimbus 7 satellite covering the time period from suitablc for data analysis. Sun contamination is a November 1985 to October 1987 were supplied by problem that occurs near sunrise and sunset with the the Nimbus 7 Project at the NASA Goddard Space WFOV radiometer because the field of view is larger Flight Center. Because of operational constraints than tile Earth's disk, and hence it views the Sun on the Nimbus 7 satellite, ERB measurements were directly. Measurements were eliminated when the not taken during the time period from April 10 to solar zenith angle at nadir (or subsatellite) point was June 23, 1986. Thus, no monthly radiation budget between 99 ° and 123 °. This was the range of angles data exist for May 1986, and data for April and used by other investigators when analyzing Nimbus 7 June 1986 are limited to the first 9 days for April ERB WFOV data (Kyle et al. 1984). With the range and the last week for June. Therefore, the monthly of solar zenith angles specified by the Nimbus 7 in- average fields for April and June 1986 should not be vestigators, it was found that another possible source treated as equivalent to monthly averages in other of error in OLR measurements, caused by a thermal years. Only OLR data are documented in this atlas. transient in SW measurements, was eliminated. The nominal duty cycle of the ERB radiometer The measurements were then time. and space- was 3 days on followed by 1 day off. When the averaged to obtain regional monthly averages, and radiometer was turned on, measurements were taken a deconvolution (resolution enhancement) technique at 4-sec intervals along the orbit track. The data have was applied to represent the radiant exitance at the been reduced in volume by averaging four consecutive TOA by a truncated series of spherical harmonics. measurements, resulting in one averaged value every The deconvolution technique takes advantage of the 16 sec. fact that spherical harmonics are the eigenfunctions Although the Nimbus 7 ERB radiometer was ini- of the measurement operator and reduces the radiant tially constrained to operate on a duty cycle of 3 days exitance field from satellite altitude to the TOA by on and 1 day off, this constraint was relaxed in dividing by the appropriate eigenvalues. All the September 1983 because of the failure of other exper- results in this atlas are based on the deconvolution iments. The ERB radiometer operated without inter- technique (Smith and Green 1981; Bess, Green, and ruption most of the time except for short periods in Smith 1981). the summer and fall of 1984 when the constraint was The governing equation from which the monthly in effect. However, the absence of a continuous daily averaged harmonic coefficients wcrc produced is data set did not seriously constrain monthly averaged WFOV data because the data wcrc smoothed in the N n averaging process over 1 month and over the large spatial area of the WFOV radiometer.